【作者】 介鸣； 黄显林；

【Author】 Jie Ming, Huang Xianlin Center for Control Theory and Guidance Technology, Harbin Institute of Technology, Harbin 150001

【机构】 哈尔滨工业大学控制理论与制导技术研究中心；

【摘要】 针对月面探测器在接近、下降、着陆阶段所要完成的精确导航任务,本文提出了一种基于月面特征点配准的月面探测器的视觉导航方法。月面探测器下降过程中,与探测器固联的CCD相机所拍摄的月面图像相对于图像数据库存在较大的旋转和缩放。本文采用了Scale Invariant Feature Transform(SIFT)算法提取月貌特征并在图像数据库中进行匹配,克服了存在较大的旋转和缩放的图像的配准问题。最后,利用月貌匹配点的地理信息,通过2D／3D位姿估计方法,估计出探测器在地理坐标系下绝对的位置和姿态。仿真结果表明该方法有效,可以应用在月面探测器在接近、下降、着陆段的导航任务中。更多还原

【Abstract】 Aimed at accomplishing lunar lander precise navigation task during the entry, descent and landing (EDL) phase,a lunar terrain registration based vision navigation method of lunar lander is proposed. During the descent phase of the lander,there exits large rotation and scale change between the images of lunar surface captured by a CCD camera mounted on the lander and the image database. The scale invariant feature transform(SIFT)method is used to select features of lunar terrain and match them in the image database, which overcomes the problem of matching images with large rotation and scale change. Finally, by using the geography information of the matched feature, absolute attitude and position in geography coordinates can be estimated by 2D/3D pose estimation. The results of simulation show that this vision navigation method is effective and can be used in the EDL navigation task.更多还原